Peroxytrifluoroacetic Acid. III. The Hydroxylation of Olefins1

period. The solution was then heated under reflux for 30 minutes and cooled. It was poured into 150 ml. of crushed ice-water mixture, and the light br...
0 downloads 0 Views 432KB Size
3472

WILLIAMD. EMMONS, ANGELOS. PAGANO AND JEREMIAH P. FREEMAN

peroxide in 100 ml. of methylene chloride was added a t room temperature 20.2 ml. (0.144 mole) of trifluoroacetic anhydride in one portion. After an induction period of one or two minutes an exothermic reaction took place causing the solution to reflux. The homogeneous mixture was then stirred a few minutes and a slurry of 9.3 g. (0.03 mole) of 2,4,6-tribromoaniline in 20 ml. of methylene chloride was added over a ten-minute period. The solution was then heated under reflux 30 minutes. I t was cooled and washed with 100 ml. of water. The turbid brown methylene chloride extract was separated and the volatile solvent was removed under reduced pressure. There was obtained as a residue 10.2 g. (100%) of yellow 2,4,6-tribromonitrobenzene, m.p. 113-115". This was recrystallized once from ethanol and melted a t 124-125' (lit. m.p. 125"). Oxidation of p-Bromoaniline to p-Bromonitrobenzene A solution of peroxytrifluoroacetic acid was prepared by addition of 5.4 ml (0.2 mole) of 90% hydrogen peroxide to 50 ml. of trifluoroacetic acid. To this solution was added fi 9 g. (0.04 mole) of p-bromoaniline in one portion. The solution was then warmed to 55' where an exothermic reaction was observed The temperature of the solution wa: allowed to increase to 85' and was maintained a t 75-85 by intermittent cooling with an ice-bath. After 15 minutes, evolution of heat had ceased, and the mixture was allowed to cool to room temperature. The solution was then poured into 200 ml. of a crushed ice-water mixture and the light tan solid which separated was collected on a filter. There was obtained 6.9 g. (85'Xa) of p-bromonitrobenzene, m.p. 122-124'. After recrystallization from ethanol t h c sample melted a t 125-126".

THE

Oxidation of p-Phenylenediamine to p-Dinitrobenzene.To 40 ml. of trifluoroacetic acid was added a t room temperature 2.2 g. (0.02 mole) of p-phenylenediamine in one portion. To this solution was added with stirring 5.4 ml. (0.2 mole) of 90% hydrogen peroxide. No evolution of heat was observed until the mixture was heated to 50'. A t this point an exothermic reaction was noted, and the temperature was kept between 60-70' for 45 minutes by intermittent cooling with an ice-bath. After the reaction was over the solution was heated a t 60' for 15 minutes. The mixture was then quenched in 200 ml. of a crushed ice-water mixture. The solid p-dinitrobenzene so obtained was collected on a filter and dried; yield 3.0 g. (89%), m.p. 160-164". The sampleowas recrystallized from ethanol once and melted a t 169-171 . Oxidation of p-Chloroaniline to p-Chloronitrobenzene with 30% Hydrogen Peroxide.-A solution of 6.4 g. of p-chloroaniline (0.05 mole) in 50 ml. of trifluoroacetic acid was heated to reflux. To this refluxing solution was added 28.3 g. (0.25 mole) of 30% hydrogen peroxide over a 30-minute period. The solution was then heated under reflux for 30 minutes and cooled. I t was poured into 150 ml. of crushed ice-water mixture, and the light brown precipitate of pchloronitrobenzene was collected on a filter and dried; yield 6.0 g. (71%), m.p. 73-75'. A sample of this material was Attempts purified by vacuum sublimation, m p. 76-78'. to raise the melting point to the literature value of 83'* by further purification were unsuccessful. ( 8 ) I Heilbron, "Dictionary of Organic Compounds University Press, New York, N Y , 1946, p. 503.

"

Oxford

HUNTSVILLE, ALABAMA

(7) K. J. P. Orton, J. Chern. Soc., 8 9 , 797 (1903).

[COXTRIBUTION FROM

Vol. 76

ROHMAND HAASCOMPANY, REDSTONE ARSENAL RESEARCH DIVISION]

Peroxytrifluoroacetic Acid.

111. The Hydroxylation of Olefins1

BY WILLIAMD. EMMONS, ANGELOS. PAGAXO AND JEREMIAH P. FREEMAN RECEIVED MARCH8, 1954 Peroxytrifluoroacetic acid has been found to be a superior reagent for olefin hydroxylation. The reagent is particularly useful for preparation of water-soluble a-glycols and for hydroxylation of negatively substituted olefins.

Peroxytrifluoroacetic acid has been found to be extremely reactive in the hydroxylation of olefins and excellent yields of the corresponding a-glycols have been obtained. Most organic peroxyacids have a t one time or another been employed for olefin hydroxylation,2 and performic acid is particularly effective in this reaction. Peroxytrifluoroacetic acid is, however, far more reactive than this reagent and in addition offers some practical advantages in synthesis of water-soluble a-glycols. Peroxytrifluoroacetic acid, prepared from trifluoroacetic anhydride and 90% hydrogen peroxide in methylene chloride, reacted almost instantaneously even a t ice-bath temperatures with all the simple olefins investigated. The products obtained were hydroxytrifluoroacetates derived from the ring opening of the intermediate epoxide with trifluoroacetic acid. These esters could, however, readily be converted to the a-glycols by m e t h a n ~ l y s i s . Ini~ tially it was found that the glycols were contaminated by high-boiling products of the cellosolve type. These were undoubtedly derived from condensation (1) T h i 5 research was carried out under Army Ordnance Contract W-O~.OZI-C)RD-~~~. ( 2 ) D. Srvern, "Organic Reactions," Vol, V I I , John Wiley and Sons, Inc., New York, S . Y . ,1YA3, p. 378. (3) E. J. Bourne, C. E M. Tatlow and J . C. T e t l d e r , J , Cizem. S u c . , 1367 (1950).

of the epoxide, initially formed in the reaction, with the hydroxytrifluoroacetate which accumulates as the reaction proceeds. Formation of these high boiling ethers was prevented, however, by increasing the effective concentration of trifluoroacetate ions present in the reaction medium; this was conRCH=CHR

CHSCOsH A

RCHCHR '0'

CF3C02H

A

RCHOHCH(0COCFa)R

&RCHOHCH(OCOCF,)R RCHOHCHROCHRCH( 0COCFs)R

veniently accomplished by addition of triethylammonium trifluoroacetate to the solvent in which the oxidation was carried out, and under these conditions pure a-glycols were obtained. In this connection it was observed that the salts of tertiary aliphatic amines and trifluoroacetic acid are extremely soluble in those chlorinated solvents which are capable of hydrogen bonding. A summary of the experimental investigations may be found in Table I. In general the oxidations were carried out by addition of a methylene chloride solution of peroxytrifluoroacetic acid to the olefin and triethylammo-

HYDROXYLATION OF OLEFINS

July 5 , 1954

3473

complex of the type indicated below. The fact that the uncatalyzed methanolysis of thetrifluoro-

TABLE I HYDKOXYLATION OF OLEFINS Yield,

Olefin

Glycol

%

Allyl chloride Allyl ethyl ether Pentene-1 Pentene-2 Hexene-1 Cyclohexene Octene-1 Dodecene-1 Tetradecene-1 Oleic acid Methyl acrylate Ethyl acrylate Methyl methacrylate

3-Chloro-1,2-propanediol

70

3-Ethoxy-1 ,%propanediol 1,a-Pentanediol 2,3-Pentanediol 1,a-Hexanediol trans-1 ,Z-Cyclohexanediol 1,2-Octanediol 1,2-Dodecanediol 1,2-Tetradecanediol 9,lO-Dihydroxystearic acid Methyl glycerate Ethyl glycerate Methyl 1,2-dihydroxyisobutyrate Ethyl 1,2-dihydroxyisobutyrate

60 77

RCH-CH2

74 80

acetate of trans-l,2-~yclohexanediol was fairly rapid, whereas that of the other glycols was relatively slow, offers some support for this hypothesis; in the latter case the formation of such an intramolecular complex is, of course, sterically impossible. 77 It is interesting to note that methanolysis of the 71 trifluoroacetate esters of methyl glycerate and 73 methyl 1,2-dihydroxyisobutyratewith methanolic hydrogen chloride was relatively unsatisfactory due Ethyl methacrylate 76 to the fact that transesterification involving the carboxyl and hydroxyl groups of the glyceric ester took place. This yielded appreciable amounts of niurn trifluoroacetate. A somewhat better grade low molecular weight polymer. This difficulty was of product was obtained if the reaction was run in solved, however, by using as a catalyst a sulfonic this sequence rather than in the inverse order. acid ion exchange resin, Amberlite IR-120, which Normally a 10% excess of the peroxyacid was em- when stirred with a refluxing solution of the triployed, and considerable care was necessary in fluoroacetate in methanol yielded the glycerate handling this very volatile reagent since i t evapo- without any polymer being formed. Under these rates rapidly from solution. The reaction times em- conditions the resin acts as a true catalyst and can ployed with peroxytrifluoroacetic acid can be made be reused any number of times. The substitution extremely short if sufficient cooling for the reaction mixture is supplied. Even the higher molecular of methanolysis for alkaline hydrolysis of the hyweight alkenes with terminal unsaturation reacted clrwy&km is cpltai&y advantagenus in g l y d synthesis since it simplifies the preparation of the almost instantaneously with peroxytrifluoroacetic acid whereas these olefins required 8 to 24 hours for very water-soluble, low molecular weight glycols hydroxylation with performic acid a t 40°.* It enormously. Tt also should be mentioned that the also has been possible to hydroxylate negatively products obtained by methanolysis were generally substituted olefins such as acrylates and methacryl- of excellent purity and boiled over a very narrow ates in good yield with peroxytrifluoroacetic acid. range. The yields listed in Table I for the low moIndeed this reagent offers a convenient laboratory lecular weight glycols are undoubtedly on the low side since they reflect the mechanical losses of two CFPCOBH distillations in most cases. It is our opinion that CHz=CHCOzR CHzOHCHOHCOzR the hydroxylation reaction itself is quantitative. CHaOH, H e The conversion of cyclohexene into trans-l,2-cyclosynthesis of glycerates and 1,2-dihydroxyisobuty- hexanediol also indicates that hydroxylation with rates. The hydroxylation of some cy,P-unsaturated peroxytrifluoroacetic acid is stereospecific and acids with performic acid has been reported but the yields trans glycols as do other peracids. yields were low and a very large excess of peracid The mechanism of olefin hydroxylation with perwas r e q ~ i r e d . ~ oxytrifluoroacetic acid may well be similar to that I n most of the experiments the isolation of the proposed for other organic peroxyacids.2 In this intermediate hydroxytrifluoroacetate by distilla- system the active reagent would be regarded as the tion was necessary in order to separate this product conjugate acid of peroxytrifluoroacetic acid. It is from triethylammonium trifluoroacetate. The ester equally prcbable, however, that perexytriflucreacealso contained some ditrifluoroacetate and possibly tic acid itself, rather than its protonated complex, some of the cy-glycol itself since trifluoroacetates is the active species which attacks the olefin. Inundergo transesterification very readily. Conse0 quently, the wide range of boiling points observed 11 for these intermediates is not surprising. The hyCFaCOOH He CFJC-0-OH droxytrifluoroacetates were in most cases converted to glycols by m e t h a n ~ l y s i s . ~It was found, how- deed, this hypothesis has some support in that the ever, that methanolic hydrogen chloride was a oxidation of olefins in the presence of trifluoromuch more satisfactory reagent for this reaction acetate ions proceeds, in a gross sense, as rapidly than methanol itself; indeed, the uncatalyzed as when no trifluoroacetate is present; addition of methanolysis reaction usually required several days triffuoroacetate ion would, of course, substantially a t reflux before i t was complete. This may pos- decrease the concentration of the protonated persibly be due to the formation of an intramolecular oxyacid complex. I n any case, however, peroxytrifluoroacetic acid has a much weaker oxygen-oxygen (4) D. Swern, G. N. Billen a n d J. T. Scanlan, THISJ O U R N A L , 68, bond than that of the known organic peracids 1504 (1946). and the reactivity of this reagent toward any ( 5 ) J. English and J. D. Gregory, ibrd., 69, 2120 (1947). 82 80 95 92 92

+

[

;i Y

]"

3474

WILLIAM

D. EMMONS, ANGEL0

s. P A G A N 0 AND JEREMIAH

nucleophilic species such as an olefin is much greater than is observed with other peroxyacids. Experimental Triethylammonium Trifluoroacetate.-To 114 g. (1 .0 mole) of trifluoroacetic acid in 100 ml. of dry ether cooled in an ice-bath was added dropwise with stirring 101 g. (1.0 mole) of triethylamine. After addition was complete, the ether was evaporated under reduced pressure to yie1,d triethylammonium trifluoroacetate as a colorless, VISCOUS liquid, yield 212 g. (99%). And. C a l d ~far ci8H&&;E3: neat. equiv., 215, Found: neut. equiv., 217 (perchloric acid in acetic acid). 2,3-Pentanediol.-To a stirred suspension of 6.0 ml. (0.22 mole) of 90% hydrogen peroxide in 50 ml. of methylene chloride cooled in an ice-bath was added 37.2 ml. (0.264 mole) of trifluoroacetic anhydride. The resulting solution was stirred ten minutes in the cold. I t was then transferred while cold to a dropping funnel equipped with a pressure equalizer tube and added dropwise over a 30-minute period to a solution of 14.0 g. (0.2 mole) of pentene-2 and 21.4 g. (0.1 mole) of triethylammonium trifluoroacetate in 50 ml. of methylene chloride. Throughout the addition, the exothermic reaction caused the solvent t o boil vigorously. After addition was complete the solution was stirred 15 minutes a t room temperature. The volatile solvents were then removed at reduced pressure. The residual colorless liquid was distilled to yield 50.5 g. of hpdtoxytrifluoroacetate, b.p. 38-60' (2.0 mm.). The triethylammonium trifluoroacetate remained in the distillation pot as residual liquid. The trifluoroacetic ester was then heated two hours under reflux with 300 ml. of 3% methanolic hydrogen chloride. The solvent was removed in vacuo, and the product was distilled t o yield- 1-5.1-g. (74%) of colorless 2,3-pentanediol,t b.p. 58-59' (0.5 mm.), V * ~ D 1.4412. 1,2-Pentanediol.-This diol was prepared by essentially the procedure described above. There was obtained 49.6 g. of hydroxytrifluoroacetate, b.p. 38-68' (3.0 mm.), which yielded after methanolysis 16.0 g. (77%) of colorless 1,2pentanediol,? b,p. 78-80°(0.3 mm.), f i 2 0 ~1.4400. 1,2-Hexanediol.-The procedure for this glycol was again similar to that described above. The intermediate trifluoroacetate ester boiled from 48-100" (1.0 mm.), yield 59.6 g., and some decomposition of the triethylammonium trifluoroacetate was observed a t the end of the distillation. Methanolysis of this ester yielded 18.9 g. (goyo) of 1,2~ hexanediol,* b.p. 96-98"(1.Omm.), n 2 0 1.4428. 1,2-Octanediol.-l,2-Octanedio! was prepared in the usual manner. There was obtained 64.9 g. of the intermediate ester, b.p. 42-80' (2.0 mm.), which yielded after methanolysis 23.3 g. (80%) of 0-glycol, b.p. 103-105' (0.5 mm.). After standing three days a t room temperature, this product crystallized t o yield a colorless, crystalline solid, m.p. 2930" (lit. m.p. 30.504). trans-l,2-Cyclohexanediol.-A cold solution of peroxytrifluoroacetic acid was prepared in the usual manner from 50 ml. of methylene chloride, 37.2 ml. (0.264 mole) of trifluoroacetic anhydride and 6.0 ml. (0.22 mole) of 90% hydrogen peroxide. This solution was added dropwise over a 20-minute period t o a solution of 16.4 g. (0.2 mole) of cyclohexene and 21.4 g. (0.1 mole) of triethylammonium trifluoroacetate in 50 ml. of methylene chloride. The exothermic reaction caused the solvent to boil vigorously throughout the addition. The solution was then stirred a t room temperature for 15 minutes and the volatile solvents were removed in U L Z C U O . The residual liquid was distilled t o yield 46.7 g. of intermediate ester, b.p. 48-108' (0.6-3.0 mm.). The distillate was colorless a t first but became yellow toward the end of the distillation. Some decomposition of the triethylammonium trifluoroacetate was also observed a t the end of the distillation. The distillate was then heated under reflux with 300 ml. of methanol for 20 hours. Evaporation of this solvent yielded 19.0 g. (82%) (6) N. A. hlilas and S. Sussman, THIS JOURNAL, 69, 2345 (1937). (7) W.E. Raufman a n d R. Adams, ibid., 46, 3040 (1923). (8) P. A. Levene and H. L. Haller, J. Biol. Chcm., 7 8 , 483 (1928).

P. FREEMAN

VOl. 76

of colorless, crystalline trans-~,2-~yclohexanediol.After one recrystallization from acetone, i t melted a t 103-104' (lit. m.p. 102-103"9). 1,2-Dodecanediol.-To 25 ml. of methylene chloride cooled in an ice-bath was added 3.0 ml. (0.11 mole) of 90% hydrogen peroxide. Trifluoroacetic anhydride, 18.6 ml. (0.13 mole), was added to the resulting mixture in one portion, and the solution so obtained was stirred five minutes in the ice-bath. The resulting solution of peroxytrifluoroacetic acid in methylene chloride was then added dropwise over a 20-minute period to a stirred solution of 16.8 g. (0.1 mole) of dodecene-1 and 10.7 g. (0.05 mole) of triethylammonium trifluoroacetate in 25 ml. of methylene chloride. After addition was complete, the solution was allowed to stand a t room temperature 15 minutes, and the volatile solvent was then removed under reduced pressure. To the residual liquid was added 150 ml. of 6% methanolic hydrogen chloride, and the resulting solution- was heated- under reflux two hours. The methanol was evaporated a t reduced pressure and the product poured into 200 ml. of hot water. An oil separated which crystallized after standing overnight at room temperature. This solid omas collected on a funnel and dried in a vacuum oven a t 50 . There rvas obtained 19.2 g. (9570) of fairly pure, colorless, crystalline 1,2-dodecanediol which melted a t 56-58' (lit. m.p. 60.5''). Methyl G1ycerate.-A solution of peroxyfrifluoroacetic acid was prepared in the usual manner from 50.8 ml. (0.36 mole) of trifluoroacetic anhydride, 8.2 ml. (0.3 mole) of 90% hydrogen peroxide and 50 ml. of ethylene dichloride. To this cold solution was added 10.7 g. (0.05 mole) of triethylammonium trifluoroacetate and 17.2 g. (0.2 mole) of methyl acrylate. The resulting solution was heated rapidly to boiling where some evolution of heat was observed. The solution was heated under reflux for two hours. I t was then cooled and the volatile solvents removed under reduced pressure. The residual liquid was distilled to yield 43.0 g. of hydroxytrifluoroacetate, b.p. 45-60' (0.5 mm. 1. This product was dissolved in 300 ml. of methanol and stirred vigorously under reflux with 50 g. of Amberlite IR-120 ion exchange resin for seven hours. The resin was then removed by filtration and evaporation of the methanol filtrate yieldei the product. After distillation there was obtained 18.5 g. (77%) of colorless methyl glycerate,'O b.p. 74-78" 0 ~ When 3ya methanolic hydrogen (0.5 mm.), ~ 2 1.4502. chloride was used for methanolysis, the yield of methyl glycerate was only 36%. Methyl 1,2-Dihydroxyisobutyrate.-To 8.2 ml. (0.3 mole) of 90% hydrogen peroxide suspended in 100 ml. of methylene chloride cooled in an ice-bath was added dropwise over a ten-minute period 50.8 ml. (0.36 mole) of trifluoroacetic anhydride. T o the resulting solution was added 21.4 g. (0.1 mole) of triethylammonium trifluoroacetate. The peroxytrifluoroacetic acid so prepared was added to a boiling solution of 20.0 g. (0.2 mole) of methyl methacrylate in 50 ml. of methylene chloride over a 30-minute period. After addition was complete, the mixture was heated under reflux for one hour. I t was then worked up in the usual manner and there was obtained 53.0 g. of hydroxytrifluoroacetate, b.p. 45-60' (0.4 mm.). This product was then heated under reflux for two hours with 300 ml. of 3% methanolic hydrogen chloride. After evaporation of the solvent, the product was distilled to yield 18.9 g. (71%) of methyl 1,2-dihydroxyD (lit. isobutyrate, b.p. 60-64' (0.1 mm.), V ~ ~1.4438 1.443811). When Amberlite IR-120 was used for methanolysis, it was necessary t o heat the solvent-resin mixture for 24 hours and a 73% yield of methyl 1,2-dihydroxyisobutyrate was obtained.

Acknowledgment.-'ih-e are indebted to Dr. R. M. Ross for several helpful suggestions and for his encouragement throughout the course of these investigations. HUNTSVILLE, ALABAMA (9) A. Roebuck and H. Adkins, Org. Syntlrcscs, 28, 35 (1948). (IO) P.Frankland and J. MacGregor, J . Chcm. Soc., 63,511 (1895). (11) J. W.E. Glattfeld and W.E. Mochel, THIS JOURNAL, 60, 1011 (1938).